In multi-node networks, multiple versions of a server may be running (e.g., in a cloud or premise deployment), so that if one of the servers goes down, services may still be provided to client devices. However, existing multi-node networks may inefficiently share data between the nodes. For example, data used to service client devices may be stored in a shared server, and the node servicing a particular client device may need to retrieve the data from the shared server. In some instances, the shared server might not have the data, and the node servicing the client device may have to spend a significant amount of time finding the data.